Medically healthy, polysubstance abusing subjects will be administered open label, IV remifentanil in a series of 2 min. infusions (50 mcg.cc, dose esccalation from 4 mcg to max of 30 mcg) using a Baxter AS40A pump by an anesthesiologist while being monitored (ECG, NIBP, O2Sat) & assessed for behavioral response in the GCRC. The investigators need to know if this selective mu opiate agonist with very short hlf-life (4 min) will be euphorogenic in this cohort. Three subjects will be studied in June 1998. If the drug is euphorogenic, the results will be used to design a full protocol to characterize the drug for use in neuroimaging. Note: The invesitagors have accomplished this by mid June. The next level: Remifental plus fMRI has been approved and one subject studied via the GCRC. COCAINE DOSE-DEPENDENT ACTIVATION BY 3T fMRI. W e have previously demonstrated the feasibility (Gollub et al, JCBFM, in press) of using fMRI to investigate the acute effects of cocaine infusion and reported the network of brain regions activated by cocaine (Breiter et al, Neuron, 1997). In this study we extend our findings using a higher field strength magnet to show dose-dependence of the cocaine effects in single subjects.